With the intensification of urbanization, the number of pedestrians in densely populated areas is showing an explosive growth trend, which brings new challenges to the technological advancement of urban traffic management, intelligent security, and other fields. Pedestrian detection in densely populated areas is an important task aimed at achieving real-time recognition and tracking of all pedestrians in a specific area. This article proposes a multi-scale occluded pedestrian detection method based on deep learning. This method can effectively detect occluded pedestrians by introducing multi-scale feature maps and occlusion detection modules. The experimental results indicate that Faster RCNN has high accuracy in detecting occluded pedestrians and can be applied to pedestrian detection tasks in practical scenarios. Deep learning algorithms can be used to effectively measure the occlusion of pedestrians. Multi-scale detection methods allow for accurate detection of pedestrians at different scales. By using deep learning algorithms, patterns with masked pedestrian features can be automatically learned and judged during the detection process.


    Access

    Check access

    Check availability in my library

    Order at Subito €


    Export, share and cite



    Title :

    Multi-Scale Occluded Pedestrian Detection Based on Deep Learning


    Contributors:


    Publication date :

    2023-10-20


    Size :

    964144 byte




    Type of media :

    Conference paper


    Type of material :

    Electronic Resource


    Language :

    English




    Feature Calibration Network for Occluded Pedestrian Detection

    Zhang, Tianliang / Ye, Qixiang / Zhang, Baochang et al. | IEEE | 2022



    Small-Scale Pedestrian Detection Based on Deep Neural Network

    Han, Bing / Wang, Yunhao / Yang, Zheng et al. | IEEE | 2020